Cost and Time-Cost Effectiveness of Multiprocessing

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

Speedup and efficiency are two measures for performance of pipelined computers. Now, these measures are used to evaluate performance of parallel algorithms for multiprocessor systems. However, to evaluate the performance of a parallel algorithm, these measures consider only the computation time and number of processors used, but do not include the number of the communication links in the system. In this paper, we define two new measures, cost effectiveness and time-cost effectiveness, for evaluating performance of a parallel algorithm for a multiprocessor system. From these two measures we define two characterization factors for multiprocessor systems. We use these new characterizing factors to analyze some well-known multiprocessor systems. It is found that for a given penalty function, every multiprocessor architecture has an optimal number of processors that produces maximum profit. If “too many” processors are used, the higher cost of the system reduces the profit obtained from the faster solution. On the other hand, if “too few” processors are used, the penalty paid for taking a longer time to obtain the solution reduces the profit.

Original languageEnglish (US)
Pages (from-to)704-712
Number of pages9
JournalIEEE Transactions on Parallel and Distributed Systems
Volume4
Issue number6
DOIs
StatePublished - Jan 1 1993

Fingerprint

Multiprocessing
Cost-effectiveness
Cost effectiveness
Parallel algorithms
Multiprocessor Systems
Profitability
Parallel Algorithms
Costs
Profit
Telecommunication links
Evaluate
Penalty Function
Multiprocessor
Penalty
Speedup

Keywords

  • Cost effectiveness
  • efficiency
  • hypercubes
  • interconnection networks
  • mesh-connected computers
  • multiprocessing
  • parallel algorithms
  • speedup

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Hardware and Architecture
  • Signal Processing
  • Electrical and Electronic Engineering
  • Theoretical Computer Science

Cite this

Cost and Time-Cost Effectiveness of Multiprocessing. / Sarkar, Dilip.

In: IEEE Transactions on Parallel and Distributed Systems, Vol. 4, No. 6, 01.01.1993, p. 704-712.

Research output: Contribution to journalArticle

@article{f14579343e6645618caa4992fff97bcd,
title = "Cost and Time-Cost Effectiveness of Multiprocessing",
abstract = "Speedup and efficiency are two measures for performance of pipelined computers. Now, these measures are used to evaluate performance of parallel algorithms for multiprocessor systems. However, to evaluate the performance of a parallel algorithm, these measures consider only the computation time and number of processors used, but do not include the number of the communication links in the system. In this paper, we define two new measures, cost effectiveness and time-cost effectiveness, for evaluating performance of a parallel algorithm for a multiprocessor system. From these two measures we define two characterization factors for multiprocessor systems. We use these new characterizing factors to analyze some well-known multiprocessor systems. It is found that for a given penalty function, every multiprocessor architecture has an optimal number of processors that produces maximum profit. If “too many” processors are used, the higher cost of the system reduces the profit obtained from the faster solution. On the other hand, if “too few” processors are used, the penalty paid for taking a longer time to obtain the solution reduces the profit.",
keywords = "Cost effectiveness, efficiency, hypercubes, interconnection networks, mesh-connected computers, multiprocessing, parallel algorithms, speedup",
author = "Dilip Sarkar",
year = "1993",
month = "1",
day = "1",
doi = "10.1109/71.242152",
language = "English (US)",
volume = "4",
pages = "704--712",
journal = "IEEE Transactions on Parallel and Distributed Systems",
issn = "1045-9219",
publisher = "IEEE Computer Society",
number = "6",

}

TY - JOUR

T1 - Cost and Time-Cost Effectiveness of Multiprocessing

AU - Sarkar, Dilip

PY - 1993/1/1

Y1 - 1993/1/1

N2 - Speedup and efficiency are two measures for performance of pipelined computers. Now, these measures are used to evaluate performance of parallel algorithms for multiprocessor systems. However, to evaluate the performance of a parallel algorithm, these measures consider only the computation time and number of processors used, but do not include the number of the communication links in the system. In this paper, we define two new measures, cost effectiveness and time-cost effectiveness, for evaluating performance of a parallel algorithm for a multiprocessor system. From these two measures we define two characterization factors for multiprocessor systems. We use these new characterizing factors to analyze some well-known multiprocessor systems. It is found that for a given penalty function, every multiprocessor architecture has an optimal number of processors that produces maximum profit. If “too many” processors are used, the higher cost of the system reduces the profit obtained from the faster solution. On the other hand, if “too few” processors are used, the penalty paid for taking a longer time to obtain the solution reduces the profit.

AB - Speedup and efficiency are two measures for performance of pipelined computers. Now, these measures are used to evaluate performance of parallel algorithms for multiprocessor systems. However, to evaluate the performance of a parallel algorithm, these measures consider only the computation time and number of processors used, but do not include the number of the communication links in the system. In this paper, we define two new measures, cost effectiveness and time-cost effectiveness, for evaluating performance of a parallel algorithm for a multiprocessor system. From these two measures we define two characterization factors for multiprocessor systems. We use these new characterizing factors to analyze some well-known multiprocessor systems. It is found that for a given penalty function, every multiprocessor architecture has an optimal number of processors that produces maximum profit. If “too many” processors are used, the higher cost of the system reduces the profit obtained from the faster solution. On the other hand, if “too few” processors are used, the penalty paid for taking a longer time to obtain the solution reduces the profit.

KW - Cost effectiveness

KW - efficiency

KW - hypercubes

KW - interconnection networks

KW - mesh-connected computers

KW - multiprocessing

KW - parallel algorithms

KW - speedup

UR - http://www.scopus.com/inward/record.url?scp=0027606576&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0027606576&partnerID=8YFLogxK

U2 - 10.1109/71.242152

DO - 10.1109/71.242152

M3 - Article

VL - 4

SP - 704

EP - 712

JO - IEEE Transactions on Parallel and Distributed Systems

JF - IEEE Transactions on Parallel and Distributed Systems

SN - 1045-9219

IS - 6

ER -